DocumentationQuick Start
Getting Started
Quick StartInstall WhiskersGrant permissionsYour first dictation
Features
Push-to-talk dictationHands-free (lock) modeAI EnhancementFile transcriptionVocabulary and snippets
Settings
HotkeysModels — local vs cloudMicrophoneAI Enhancement settingsDictionaryAPI Keys panelFile Transcription panel
API Key Setup
Setting up your Groq API keySetting up your Deepgram API keySetting up your OpenAI API key
Troubleshooting
Common issues and fixes
Quick Start·2 minutes

Set up Whiskers in 2 minutes

The shortest path from download to your first transcription. If you want more depth on any step, follow the link at the bottom of each section.

Install Whiskers

Download the latest .dmg from the Whiskers website, open it, and drag the app to your Applications folder. On first launch, right-click the app icon and choose Open to approve the unsigned binary.

Mounted Whiskers .dmg in Finder, with the Whiskers app icon being dragged into the Applications shortcut

Grant two permissions

The onboarding flow walks you through both. You'll need:

  • Microphone — so Whiskers can hear you.
  • Accessibility — so Whiskers can detect your hotkey from any app and paste the transcript where your cursor is.
System Settings → Privacy & Security → Accessibility list, with the Whiskers toggle being turned on

Screen Recording permission is optional and only used for Screen Awareness (text-only context to improve accuracy).

Pick a transcription model

On first launch, choose a model in the onboarding step. For most users:

  • English-only? Pick Parakeet V2 (~400 MB) — it's the fastest local model.
  • Other languages? Pick Parakeet V3 (~600 MB, 25 languages).
  • Don't want a local download? Skip the model step and add a cloud API key later from Settings → Models → Cloud API.
Onboarding model picker showing Parakeet V2, Parakeet V3, and Whisper Large v3 cards with Download buttons

Models download in the background; the menu bar icon shows progress.

Dictate your first sentence

Click into any text field — Notes, Slack, Mail, your browser — anywhere a cursor blinks.

  1. Hold the Fn key.
  2. Speak.
  3. Release.
A Slack compose box with the cursor inside, and the Whiskers HUD floating above showing a live waveform and a 00:04 timer

The transcript pastes itself where your cursor is. That's it.

Where to go from here

Getting Started →Deeper setup: install, permissions, first dictation.Features →AI enhancement, file transcription, custom vocabulary.Settings →Hotkeys, models, microphone, and app behavior.Troubleshooting →Mic not working? Hotkey ignored? Start here.